search

spinicist / QUIT

A set of tools for processing Quantitative MR Images
Mozilla Public License 2.0
58 stars 20 forks source link

Clarifications on input structure for most functions #38

Open gdevenyi opened 4 months ago

gdevenyi commented 4 months ago

When working with our various quantitative imaging components, we follow a BIDSified format where all our inputs are separate files. It appears various functions in QUIT require images with stacked-dimensionality of some kind to store multiple image inputs for processing (ex #37)

Could you please provide some additional documentation on the expected input format for the various functions? Thanks!

spinicist commented 4 months ago

All the QUIT commands require the main inputs to be "stacked" along the 4th / time dimension in the input nifti file. My own qMRI sequences on GE write out DICOMs that will convert to the corresponding niftis, and I find the format to be far more convenient to work with than separate files. For instance, CEST experiments usually have >20 offset frequencies, it's much easier to have one file than 20 separate ones. It's also the convention for diffusion imaging.

This is one of several reasons I have ignored BIDS as much as possible.

gdevenyi commented 2 months ago

Can you please enhance the documentation regarding the expected order of these stacked 4D files where possible? Thanks.

spinicist commented 2 months ago

What documentation do you want to see and where do you want it to go?

I do not mean to be obtuse - as I started working with fsl, and this is the way all the fsl tools work, I've taken this format to be self-evident.